Hi,

I like to program microcontrollers. I know drones, but I haven't built one.

I don't want a racing drone, I'm interested in DIY using Arduino as the controller, because I want the programming part to be educational.

The parts I think are necessary to be specialized to the drone build are:
1. Motors: need to be powerful enough to lift a bit heavy frame, battery and all the rest components
2. ESC: I think these are necessary to balance the power going to the motors so I can have smooth motor control
3. Battery: It should be powerful enough to lift the drone
4. Transmitter/Receiver: Anything that works
5. Camera: Not interested in it right now
6. Fans: Which is better, 2 or 3 blades ?

But I want the rest to be custom parts and not specialized; like:
1. Frame: I want to do either custom or specialized one, I want to test the system in both frame types
2. Controller: I want to use hobby control boards; like, Arduino, Maple mini, Raspberry pi, ESP32 ... etc.


Now what I want to know is which type of ESC, battery and transmitter/receiver to choose ?

1. There are many types of ESCs, 20A, 30A, ... etc. What number to choose and which brand ?
2. Also the battery, should I get 2S, 3S, ... etc. ?

Arduino, Maple mini, Raspberry pi, ESP32????? You will spend more time porting the existing STM32 bit firmware to the new hardware platform and spend more time and money integrating the sensors you will need.

Frame: Do not even dare 3d print frames because they will not last for long. Not to mention the lack of stiffness or rigidity. Unless you build a light toothpick.

Folks in the FPV world have refined these ideas so several years now and so if you are trying to re-invest the wheel then you are in reality playing with too many variables.

So you need total weight first. That will lead to what motors and PDB/ESC you need, the FC I guess will be an ArduPilot APM. Frame and the rest is up to you.

Or you could buy a HolyBro X500 development kit, which would be easier and at least give you all the basics minus the Lipos.

First of all, if you really want to build a drone based off of an arduino, raspberry pi, etc- cool! I will happily follow along.
I originally thought it’d be cool to build a drone about over a decade ago- I reached out to someone to try to figure out what I needed to do, thinking I could base it off of an arduino- and I was basically told that it was not how things were done anymore. That turned me off enough to not bother looking into drones until 10 years later.

It can be done. But, you’re looking at tech that is not purposely designed for the task. That creates some limitations and problems, but they can be overcome. If you do a search for “arduino quadcopter” on google, you’ll find a number of projects and guides.

It sounds like you’re more interested in the programming than anything else? If this is the case, as Pathfinder said, you might be better off with a drone specific development board.

https://www.st.com/en/evaluation-tools/s...001v1.html
^this can run brushed motors, or brushless motors if you add electronic speed control. Modern quadcopters/drones typically use brushless DC (bldc) motors, but brushed motors will make life easier if you really want to program from scratch.

Here’s some brushless (bldc) esc reference designs:
https://www.mouser.com/datasheet/2/389/e...107609.pdf
https://www.ti.com/tool/TIDA-00916#tech-docs

Here’s a link to a guy who built his own flight controller from scratch:
https://www.flying-rabbit-fpv.com/2020/1...ontroller/

I can’t help you with the programming- that is *way* beyond my skill set.

To answer your specific questions, this will be ABC, A: first set of questions, etc.

A:
1- motors, 2- ESC, 3- battery, 6- fans (we call these propellers):
All of these are interrelated and dependent upon the AUW (All Up Weight) of the drone. Because this will be a completely scratch built drone, we need a lot more information to determine or guess the weight. Without having an idea of the size and weight, we can’t determine the minimum amount of thrust. Bare minimum we’d need a 2:1 thrust to weight ratio, but that is bare minimum- it wouldn’t fly well because it wouldn’t be able to make enough thrust to do anything other than very sluggish/slow movements. We can’t determine the necessary thrust without knowing the weight. We can’t know what size motor can handle the required prop size and pitch to create the needed thrust, we can’t determine how much amperage the motor and propeller will require the esc to be able to handle, and we can’t determine the appropriate size battery to supply that.
6a: the number of blades, as well as the pitch of the blades, will depend on how much thrust is needed, how much amperage and voltage the motors and esc’s can handle, and how much the battery is capable of supplying.

ESC’s are rated for amperage (and voltage based off of battery cell count… ) Motors are rated by kv- kv rating is 1000rpm per volt. There’s a great thread here about how prop size, motor kv, and battery cell count relate to current. There’s a lot of information that I’m still digesting. (Thanks V-22 for linking to this in another thread!)
https://www.rcgroups.com/forums/showthre...mple-table

B:
1- frame: if you want to do a custom frame with a custom arduino or raspberry pi or whatnot controller, I’d suggest something large. 5” bare minimum (that is 5” propeller size, not motor to motor). You will likely need to go much larger to fit your boards and lift the weight. Maybe 6”, 7”, or above?
2- controller: see above. If the programming is what you’re interested in, please be aware that both Betaflight and iNav are open source. Most of us are using one of these, there is nothing stopping you from modifying either of them to fit your needs. If you really want to go the arduino or pi route, Betaflight can run on pi, so maybe at least take a look at the code. This is not something I can help you with. I am not a programmer.

C:
Can’t answer any of these questions without the information from A.

The link below is a great tool for playing with numbers to see what size motors, propellers, and battery cell count and size will work for various drone sizes and weights. I use it often, it is worth the $7.99 for the full version to be able to plug in custom data.
https://www.ecalc.ch/xcoptercalc.php

Oscar has a good tutorial on how to pick parts here:
https://oscarliang.com/quadcopter-motor-propeller/

Thank you guys so much for the answers.

(01-Sep-2022, 02:02 AM)kafie1980 Wrote: Arduino, Maple mini, Raspberry pi, ESP32????? You will spend more time porting the existing STM32 bit firmware to the new hardware platform and spend more time and money integrating the sensors you will need.

Basically the model I want to work on is for educational purposes not for competition. Also it's not meant to be robust.

It's for educational demonstration, and to be exposed for explanation of the basics of how a drone basically works.

There are people on YouTube who do DIY drones which is exactly what I want to do.

I want to do multiple versions of these drones:
1. a DIY version which has custom parts and programming
2. a professional version which has the knows required parts that drone professional people do

Quote:Frame: Do not even dare 3d print frames because they will not last for long. Not to mention the lack of stiffness or rigidity. Unless you build a light toothpick.

3d print would be ok because it should be tested indoors, it's for students to work on and do experiments.
I'm also considering like the people who do DIY models on YouTube to use light wood or acrylic ... etc.

I advised the student who want to do the project to use either a robust material or buy a good frame from Aliexpress.

Even I want the student to try and design his own frame, just for experimenting.

Quote:Folks in the FPV world have refined these ideas so several years now and so if you are trying to re-invest the wheel then you are in reality playing with too many variables.

Yes I agree with you and I understand it's a sensitive variables to run a drone, even the custom drone model might face problems and challenges which is the goal of the student project, otherwise the college would ask the student to buy a complete professional kit and launch it and that's it. But this is not the goal of the course.

You know I would love to make a balsa wood frame. Just for the sake of doing it. It would probably not be the most solid or aerodynamic build, but it would probably fly. Even a thin plywood would probably work, like 1-2mm ply to make a 2" diamond frame, then cut a custom design into the inner area.

Now trying to make it fly using a pure arduino or RPi approach, that could be very interesting too. I actually wondered about it when i first got into this, but decided it might be easier to try and replicate the actual radio end of it, so instead of a radio sending packets to the receiver, a RPi would do it and with GPS and LiDAR/FLiR sensors, it would allow a degree of pre-programmed autonomous flight. Kind of like give it a grid reference and tell it to fly an expanding grid while mapping the ground using LiDAR and FLiR. Useful for finding things and people in forest cover, or tracking fires/hotspots.

I really hope you and your students have fun with your projects, r1s8k.

(03-Sep-2022, 10:33 PM)Pathfinder075 Wrote: You know I would love to make a balsa wood frame. Just for the sake of doing it. It would probably not be the most solid or aerodynamic build, but it would probably fly. Even a thin plywood would probably work, like 1-2mm ply to make a 2" diamond frame, then cut a custom design into the inner area.

OK, let say, if the weight calculations would be rough to guess.

What a good motors and lipo would get a normal custom drone to fly ?

I found these kits on Aliexpress, please tell me if the motors included are powerful enough:

https://www.aliexpress.com/item/40002837...EpD4z&mp=1

or

https://www.aliexpress.com/item/10050012...EpD4z&mp=1

with "2205 30A 5045 4PCS" kit.


Another question:
What about the power board ? is it necessary with the ESCs ? or I have to buy one ?

Quote:Now trying to make it fly using a pure arduino or RPi approach, that could be very interesting too. I actually wondered about it when i first got into this, but decided it might be easier to try and replicate the actual radio end of it, so instead of a radio sending packets to the receiver, a RPi would do it and with GPS and LiDAR/FLiR sensors, it would allow a degree of pre-programmed autonomous flight. Kind of like give it a grid reference and tell it to fly an expanding grid while mapping the ground using LiDAR and FLiR. Useful for finding things and people in forest cover, or tracking fires/hotspots.

I really hope you and your students have fun with your projects, r1s8k.

That's absolutely interesting ! I will definitely ask you later about the application features of using "GPS and LiDAR/FLiR sensors".

Just after I get it to work.

Depends on the weight of the frame and the type of flying you intend to do. i don't know what the exact formula is for ideal thrust to weight. I assume it to be 4:1 or 5:1, so say i build a 50g quad, I would want 200-250g of thrust, which means it should fly on 30-40% power. So if you add up all the parts for your quad build, you will get a number, times it by 4 or 5, then go and find some motors that generate that much thrust at 100%, from there you will know how many amps will be used to generate said thrust, so you now know what size ESC you need. Most heavy lifter motors will need 3-6S. Smaller quads can run on 2S. I don't think what you are planning on doing is possible on whoop sized quads, so you won't be using 1S, nor probably 2S.

I couldn't really pass judgement on those motors. I tend to avoid Ali. Also i don't fly anything that big myself, so I would be a poor choice to advise you on motors that big. But you probably will need a big ESC. Some of the motors of that size can pull up to 50A. The biggest motors I have are 1202's.